diyAudio (
-   Planars & Exotics (
-   -   distortion From ESL at low level (

WrineX 29th December 2012 07:06 PM

distortion From ESL at low level
1 Attachment(s)
i noticed that i already a looong time ago posted questions about the same subject. distortion at really low listening levels , with piano or other sound that decay, somewher ein the mid region. funny thing is this is a new pair of diy speakers contructed from Solosound panels. i have a cap on the tranny of 10 uF to start rolling of at about 500hz. i changed it with a good audin cap with the same result distortion at verry low volume. (low volume is at the amplifier not computer before someone asks)

the distortioin is not noticable when playing loud.

when i remove the cap completlely the distortion is gone. someone has an idea ? cant play them without any cap ofcourse

Here are the measurements.

Bleu Line is esl without cap red is with an audin cap of 15uF

upper line is freq response bottom line is distortion (THD) look how nice the distortion goes up at the frequency the cap kicks in, also funny is that the distortion is a few DB louder overal, i only changed the cap nothing else

this measurement is taken up close so i could do it at low volume. the Hump at 200 is resonance freq of the panel.

also the impulse repsonse changed quit a bit. the blue one looks beter overal

geraldfryjr 30th December 2012 12:10 AM

What does the signal look like coming out of the amp?

It could be that the amp is having a hard time with such a high value of capacitance on its output and can't supply enough current to make up the difference.

I had found this was happening with my setup when I was learning and testing about transformers,

Extra reactances,core saturations and Transformer resonances cause a low impedance load to the amplifier and cause more of a demand on it.
If it can't supply the current then you will get such distortions as in those pictures in the above links.

Core saturation's and Resonant frequency's provide to the amplifier what looks like a dead short too it and is why those ranges must be avoided.

Most of these pictures show the switching of the transistors and the waveforms represent the current of the signal and in one picture it shows the representation of both the current and the voltage at the same time.
Notice the 90 degree shift due to the capacitive reactance.

That is my best guess as I don't Know the full details of your setup.

I hope this helps you.

jer :)

geraldfryjr 30th December 2012 12:24 AM

Or also it could be that you are hearing the resonance of the panel too.

But, Being that it is gone after removing the capacitor it seems more of a reactive load problem with the amplifier itself.

jer :)

Calvin 30th December 2012 09:21 AM


maybe You could do an impedance measurement? This would also clear up as to which extend the serious HF-peak is due to the mic or the transformer-panel combo.
How does THD behave with a series resistor between amp and ESL?
Like Gerald I also assume rather amplifier drive probs. A Resistor of 1-5Ohms would dampen the electrical series LC-resonance and reduce loading on the amp.


WrineX 30th December 2012 10:05 AM

thanks, i did try a resistor , but 50 i wont have any upper frequency left in my panel :)

just a series resistor before the cap i presume ? or after and then recalculate cap needed.

ill make a new measurement set to see what happens with a resistor, in series. it will tame the upper hump i believe. but i think its not gone solve the distortion.

Impedance measurement i would love to make but i have never tried it, it was with a reference resistor i belief ? i have to look it up again.

By the way , how can it be hard on the amp on low voltage and no problem at all on higher levels ? also its stable to around 2 ohm should nog be that hard. i could change amp to a class a amp just to see if something changes.

WrineX 30th December 2012 12:19 PM

2 Attachment(s)
well here are the measurements up close with and without cap and a resistor of 1.2 ohm in series wth the transformer. picture one

if i increase this number for instance 18 ohm spl goes down overal and mostly above 10 k as expected. weird thing is even without cap and series resistor of 18 ohm distortion s heard in the lower regions. picture 2

WrineX 30th December 2012 01:14 PM

heres a recording of the diference with cap and without. dont mind me not finding right words :)

geraldfryjr 30th December 2012 05:42 PM

Have you tried a different cap or different values.

By looking at the waveform in Audacity, it seems that the signal is having a hard time trying to swing positive and seems to show some sign of crossover distortions.
It is hard to tell because your record level is a bit low.
It almost sounds as if there is hum being introduced.

Do you have any resistance between the bias supply output and the Diaphragm?

If even without the cap and a resistor you hear such distortion?

I added as much as 47uf to the primary side of my setup and I never had any such distortions.

What transformers are you using?

Are taking the bias supply voltage out and ground common to only one bank of the capacitors (same side)?
jer :)

WrineX 30th December 2012 06:41 PM

could it be lack of resistance at the bias out ?because i know the pannels have pretty low resistance coating, at least for nowadays. i must admit i took a few steps back on the cascade to get a lower bias, so im skipping the resistor. although because i thought maybe it has sometrhing to do with it i tried to connect it to the original bias out with the resistor and it did not help. but maybe theres not enought resistance there as well.

yes the 0 of the transformer is going to the bias.

The sound is indeed sounding like it has to overcome something before it really plays. like a barrier around the crossover frequency, above it ,its fine i believe.

geraldfryjr 30th December 2012 07:33 PM

Did you take an odd number of steps back on the cascade or or an even number?

If it is an Odd number you must switch your ground to the esl to the other side.

The same side the output is taken from or you will get hum.


All times are GMT. The time now is 01:12 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio